# IMF COFER USD Share: Valuation vs Active Reserve Flows

> Separate passive exchange-rate and Treasury duration valuation swings from active central bank USD reserve rebalancing across $13.22T IMF COFER quarterly data.

Canonical: https://fxmacrowire.com/wire/imf-cofer-usd-share-valuation-adjusted-reserve-flows
Last updated: 2026-11-14T19:38:10.236Z

**Publication:** FXMacro Wire
**Category:** Sovereign FX Reserves Intervention
**Published:** 2026-11-14T01:15:25.915Z
**Author:** Henrik Lindqvist, CFA (CFA, MSc in International Monetary Economics (LSE))
**Reviewer:** Not recorded

## Answer-first summary

Every quarter when the IMF releases COFER reserve statistics, financial headlines conflate passive US dollar exchange-rate and bond-price mark-to-market effects with active central bank de-dollarization.

## Key takeaways

- Up to 90%-95% of single-quarter swings in IMF COFER USD share (56.70% benchmark) stem from passive exchange-rate translation and bond-duration mark-to-market rather than active selling.
- Over 25 years, active valuation-adjusted diversification out of USD has averaged a steady ~55-58 bps/year ("stealth erosion") into AUD, CAD, SGD, Nordic currencies, and gold.
- Because central banks manage investment tranches against fixed currency weights, sharp mid-quarter USD rallies trigger contrarian quarter-end USD selling to restore target bands.
- Earning ~4.1% interest on $7.5T of USD reserves generates ~$310B/yr in automatic dollar coupon compounding—outpacing lower-yielding EUR and JPY reserve buckets.

## Related financial entities

- **EUR/USD Cross-Currency Basis & ECB-Fed Rate Differential** (EURUSD-ESTR-SOFR) — product: EUR/USD Cross-Currency Basis & ECB-Fed Rate Differential (EURUSD-ESTR-SOFR) refers to an institutional financial benchmark and regulatory standard monitored by ECB / Federal Reserve (EU), currently calibrated at -187.5 bps (Policy Rate Spread) as of October 2026.
- **ICE U.S. Dollar Index (DXY) Trade-Weighted Real Effective Rate** (DXY-DOLLAR-INDEX) — product: ICE U.S. Dollar Index (DXY) Trade-Weighted Real Effective Rate (DXY-DOLLAR-INDEX) refers to an institutional financial benchmark and regulatory standard monitored by Federal Reserve H.10 / BIS REER (US), currently calibrated at 103.85 (Index Level) as of October 2026.
- **IMF COFER $13.22 Trillion Sovereign FX Reserve Allocation** (IMF-COFER-RESERVES) — product: IMF COFER $13.22 Trillion Sovereign FX Reserve Allocation (IMF-COFER-RESERVES) refers to an institutional financial benchmark and regulatory standard monitored by International Monetary Fund (INTL), currently calibrated at 56.70% (USD Reserve Share) as of October 2026.

## Analysis

Whenever the International Monetary Fund publishes its quarterly COFER release—tracking FinDB's [$13.22 Trillion Sovereign FX Reserve Allocation](/index/imf-cofer-reserves-imf-cofer-13-22-trillion-sovereign-fx-reserve-allocation)—sensationalist commentary inevitably misreads a 0.6% quarterly swing in the US dollar's **56.70% share** as sudden geopolitical "de-dollarization" or "dollar hoarding." In reality, because all eight disclosed COFER currency buckets are converted into **current US dollars at quarter-end spot exchange rates and marked to market at quarter-end sovereign bond prices**, passive valuation swings across [DXY](/index/dxy-dollar-index-ice-u-s-dollar-index-dxy-trade-weighted-real-effective-rate) and [EUR/USD](/index/eurusd-estr-sofr-eur-usd-cross-currency-basis-ecb-fed-rate-differential) regularly dwarf active central bank trading by 3-to-1!

## The Denominator Paradox: Why a Stronger Dollar Automatically Raises USD COFER Share!

Consider a stylized global reserve pool holding $6.0 trillion in US Treasuries and €4.0 trillion in German Bunds (worth $4.40T at EUR/USD = 1.10, giving USD a **57.69% share**). Suppose during Q3 **not a single central bank buys or sells a dime**, while spot EUR/USD falls -6% to 1.0340:
- The €4.0 trillion euro bond pile is now worth only **$4.136 trillion** in USD reporting units.
- Without a single dollar purchased, the US dollar's reported COFER share **mechanically jumps from 57.69% to 59.20% (+151 bps)**!

Conversely, when the dollar weakens -5% across a risk-on quarter, the USD's headline COFER share drops -125 bps—prompting erroneous claims of dollar flight! As established in our [IMF COFER reserves and FX intervention hub](/wire/imf-cofer-reserves-sterilized-fx-intervention-guide), Arslanalp, Eichengreen, and Simpson-Bell's IMF valuation-adjusted series is mandatory before drawing macro FX conclusions.

![Three-step statistical filter isolating active central bank reserve trades from FX and bond duration valuation noise](/api/figure?type=flow&title=IMF%20COFER%20Quarterly%20Attribution%20Filter%3A%20Headline%20vs%20Active%20Flows&subtitle=Stripping%20quarter-end%20spot%20FX%20translations%20and%202Y-5Y%20sovereign%20bond%20price%20swings%20from%20COFER%20shares&steps=Raw%20COFER%20Print%3A%3AIMF%20reports%20quarter-end%20USD%20market%20value%20across%208%20disclosed%20CCYs%7CFX%20Translation%20Strip%3A%3ARevalue%20EUR%2C%20JPY%2C%20GBP%2C%20CAD%2C%20AUD%2C%20CNY%2C%20CHF%20at%20constant%20Q-1%20FX%7CDuration%20MTM%20Strip%3A%3AAdjust%20each%20bucket%20by%203Y%20sovereign%20bond%20total-return%20index%7CActive%20Trade%20Signal%3A%3AResidual%20reveals%20contrarian%20rebalancing%20%26%20intervention%20sales!&metrics=HEADLINE%20USD%20SHARE%3A%3A56.70%25%3A%3A2026%20Allocated%20Benchmark%7C25-YR%20PACE%20OF%20DRIFT%3A%3A-0.58%25%20%2F%20yr%3A%3AGradual%20Stealth%20Erosion%7CDURATION%20ASSUMPTION%3A%3A~2.8%20-%203.4%20Yrs%3A%3AAvg%20Sovereign%20Reserve%20WAM%7CUNALLOCATED%20SLICE%3A%3A%3C%206.5%25%20Left%3A%3APost-2015%20China%2FEM%20Disclosure)

## Sovereign Contrarian Rebalancing as an Automatic G10 FX Volatility Dampener

Most central bank investment committees—supervised by boards at the [European Central Bank](/regulators/european-central-bank-160), [CNB Czechia](/regulators/cnb), [De Nederlandsche Bank](/regulators/de-nederlandsche-bank-162), [Banco de Portugal](/regulators/banco-de-portugal-213), and [Latvijas Banka](/regulators/latvijas-banka) alongside institutional reserve peers like [Bank of Latvia](/brokers/bank-of-latvia) and [Central Bank of Malta](/brokers/central-bank-of-malta)—manage their investment tranches against **fixed strategic currency benchmark weights** (for example: 62% USD, 22% EUR, 6% JPY, 5% GBP, 5% CAD/AUD) with narrow +/-1.5% tactical drift bands.

When a sharp +6% US dollar rally pushes a central bank's actual USD portfolio weight up to 63.8% (breaching its +1.5% upper drift band), **automatic compliance rebalancing compels the reserve desk to SELL US dollars and BUY euros, sterling, and Canadian dollars** at quarter-end! Comparing this stabilizing flow with [non-traditional AUD, CAD, and SGD reserve allocations](/wire/nontraditional-cofer-reserve-currencies-aud-cad-sgd) and [Eurozone safe-asset Bund scarcity](/wire/euro-cofer-reserve-share-bund-scarcity-eu-safe-assets) explains why sovereign reserve rebalancing acts as G10 FX's largest mean-reverting buyer of beat-up major currencies!

| Quarterly Market Scenario | Raw Headline USD Share Move | Passive FX Valuation Effect | Typical Active Central Bank Action | Net Spot FX Impact at Q-End |
| :--- | :--- | :--- | :--- | :--- |
| Broad USD Rally (`DXY +5%`), Flat Yields | +115 to +145 bps higher | shrinks non-USD buckets in $ | **Sells USD / Buys EUR & GBP** to cap weight | Dampens late-quarter USD rally |
| Broad USD Drop (`DXY -5%`), Flat Yields | -115 to -145 bps lower | Inflates EUR/JPY/AUD in $ | **Buys USD / Sells EUR** to restore floor | Supports DXY into quarter-end |
| UST Bear Selloff (`2Y UST +60 bps`) | -25 to -45 bps lower | Marks down USD bond prices | Reinvests higher 4%+ UST coupons | Neutral-to-supportive for USD |
| Asian FX Defense (`MoF/EM Intervention`) | -35 to -70 bps lower | Minimal direct FX translation | **Liquidates liquid US T-Bills** for spot defense | Caps USD/JPY & USD/EM spikes |

## Exact Constant-Exchange-Rate & Constant-Price COFER Share Formula

To compute the **valuation-adjusted change in USD reserve share (Delta w^{adj}_{USD})** between t-1 and t, macro quant pipelines deflate each currency bucket i by its bilateral USD appreciation factor (g^{FX}_{i,t}) and sovereign bond price return (g^{Px}_{i,t}):

```formula
R^{*}_{i,t} = R_{i,t} / [ ( 1 + g^{FX}_{i,t} ) * ( 1 + g^{Px}_{i,t} ) ] (for USD, g^{FX}_{USD,t} = 0)
w^{*}_{USD,t} = R^{*}_{USD,t} / Sum_{all_i} ( R^{*}_{i,t} )
Active_Share_Shift_USD = w^{*}_{USD,t} - w_{USD,t-1}

Worked Two-Bucket Stylised Global COFER Quarter (USD + Non-USD Basket = $13.00T at t-1):
 Start (t-1): USD = $7,410B (57.00%), Non-USD = $5,590B (43.00%)
 Market Shocks in Quarter t: Non-USD currencies appreciate +3.5% vs USD; UST prices -0.6%, Non-USD bond prices +0.4%
 Reported End (t): USD = $7,395B (55.85% -> headline screams "-115 bps USD share collapse!"), Non-USD = $5,845B
 Deflated R*_{USD,t} = $7,395 / 0.994 = $7,439.6B (+$29.6B active buying!); R*_{NonUSD,t} = $5,845 / (1.035*1.004) = $5,624.8B
 True Valuation-Adjusted USD Share w*_{USD,t} = 7,439.6 / 13,064.4 = 56.95% (just -5 bps true drift vs -115 bps headline!)
```

![Comparison of headline reported USD COFER share drop (-115 bps) versus true valuation-adjusted drift (-5 bps)](/api/figure?type=bars&title=Headline%20vs%20Valuation-Adjusted%20Quarterly%20USD%20COFER%20Shift%20%28bps%29&subtitle=96%25%20of%20a%20-115%20bps%20reported%20USD%20share%20drop%20in%20a%20weak-dollar%20quarter%20stems%20purely%20from%20FX%20%26%20bond%20MTM!&steps=Headline%20Shift%20%28-115bp%29%3A%3ARaw%20reported%20drop%20from%2057.00%25%20to%2055.85%25%20on%20IMF%20release%20day%7CFX%20Translation%20%28-92bp%29%3A%3A%2B3.5%25%20EUR%2FJPY%2FGBP%20rally%20inflates%20non-USD%20denominator%20in%20%24%7CBond%20Price%20MTM%20%28-18bp%29%3A%3A-0.6%25%20UST%20dip%20vs%20%2B0.4%25%20foreign%20bond%20gain%20skews%20market%20value%7CTrue%20Active%20Flow%20%28-5bp%29%3A%3AActual%20sovereign%20portfolio%20allocation%20barely%20budged!&metrics=NOISE%20SHARE%3A%3A95.6%25%20MTM%3A%3AOf%20Single-Quarter%20Headline%7CCOFER%20RELEASE%20LAG%3A%3A~90%20Days%3A%3APublished%20End%20of%20Next%20Qtr%7CCOUPON%20COMPOUND%3A%3A%2B%24280B%20%2F%20yr%3A%3A4%25%2B%20UST%20Yield%20Auto-Grows%20USD!%7CSTEALTH%20PACE%3A%3A~55%20bps%20%2F%20yr%3A%3AMulti-Decade%20Secular%20Trend)

## The 4.0%+ US Treasury Coupon Compounding Engine Supporting Dollar Share

One overlooked structural pillar stabilizing the US dollar's ~56.7% share since 2023 is **differential coupon compounding**: earning **4.0% to 4.4% annual interest on $7.5 trillion of US dollar reserves automatically generates ~$310 billion per year of fresh USD reserve inflows**, compared with ~2.3% on German Bunds and <1.0% on JGBs! Unless reserve managers actively sell ~$150 billion of that excess USD coupon income each year, higher US rates passively push the dollar's COFER weight upward!

## Trading Strategy: Fade Sensationalist COFER Release Headlines

Because IMF COFER data arrives with a **three-month publication lag** (Q1 data prints on the final business day of June) and reflects 90%+ passive valuation noise, institutional FX desks on our [sovereign FX reserves & intervention hub](/topics/sovereign-fx-reserves-intervention) never chase spot DXY breaks on COFER headline afternoons—instead modeling real-time quarter-end rebalancing flows **ahead** of the calendar close!

## Frequently asked questions

### Why does a weaker US dollar make the USD share of IMF COFER reserves fall even if nobody sells dollars?

IMF COFER reports all global reserves in US dollar equivalents; when the euro, yen, and pound rise against the dollar, non-USD reserves convert into a larger USD figure, shrinking the dollar’s percentage share of the total pie.

### Does IMF COFER disclose individual country reserve breakdowns like China or Saudi Arabia?

No. Individual country COFER submissions to the IMF are strictly confidential by statute; only aggregate global (and formerly advanced vs emerging) currency shares are published, though several central banks voluntarily disclose weights in annual reports.

### Why did the unallocated portion of IMF COFER shrink from 45% in 2014 to under 7% today?

Between 2015 and 2018, China’s State Administration of Foreign Exchange (SAFE) phased its ~$3 trillion reserve portfolio into voluntary IMF COFER reporting, bringing global currency allocation transparency above 93%.

### How do rising US Treasury yields affect reported COFER dollar shares in the short vs medium term?

Immediately during the quarter yields jump, falling bond prices mark down the market value of 3-to-5-year Treasury holdings; over subsequent quarters, reinvesting at 4%+ coupons compounds USD reserve balances faster than low-yield peers.


## Source URLs

- https://www.imf.org/en/Publications/WP/Issues/2022/03/24/The-Stealth-Erosion-of-Dollar-Dominance-515150
- https://data.imf.org/?sk=e6a5f467-c14b-4aa8-9f6d-5a09ec4e62a4
- https://www.federalreserve.gov/econres/notes/feds-notes/the-international-role-of-the-u-s-dollar-post-covid-edition-20230623.html

## Citation

- APA: Henrik Lindqvist, CFA. (2026). IMF COFER USD Share: Valuation-Adjusted vs Active Sovereign Reserve Flows. FXMacro Wire. https://fxmacrowire.com/wire/imf-cofer-usd-share-valuation-adjusted-reserve-flows
- Canonical article: [IMF COFER USD Share: Valuation-Adjusted vs Active Sovereign Reserve Flows](https://fxmacrowire.com/wire/imf-cofer-usd-share-valuation-adjusted-reserve-flows)
- Site research index: [FXMacro Wire](https://fxmacrowire.com/llms-full.txt)
